All indicators

EWMA Volatility

By LuxAlgoJun 16, 2026

Static chart image

EWMA Volatility is the definitive clean build of EWMA volatility estimation: each bar the variance estimate keeps a lambda share of yesterday's estimate and takes the rest from yesterday's squared log return, so a shock enters immediately and decays geometrically instead of dropping out of a window. The reading is annualized by default, filled with a gradient anchored at zero, and paired with a slower equal-weight benchmark on the same returns.

How to Trade the EWMA Volatility?

  • EWMA above the Equal-Weight line: turbulence is recent and building; the crossover fires the Turbulence Building alert.
  • EWMA back below: the shock is decaying — the market is settling.
  • Percentile rank above 80: an expansion regime by the instrument's own history.
  • Rank below 20: compression — a quiet stretch that often precedes range expansion.

The estimate sizes movement and carries no direction, like everything in the volatility family; use it for expectations and risk, not bias.

EWMA Volatility Settings

  • Lambda (Decay Factor) (default 0.94): the RiskMetrics daily convention. Higher decays slower and smoother; lower chases shocks quickly but runs noisier.
  • Source (default close): series the close-to-close log returns come from.
  • Annualize (default on) and Bars Per Year (default 252): scales the reading for cross-instrument comparison.
  • Equal-Weight Comparison (default on) and Comparison Window (default 60): the fixed-window benchmark behind the fast-versus-slow read.
  • Percentile Lookback (default 100) with High Percentile (80) and Low Percentile (20): rank thresholds driving the expansion and compression alerts.
  • Show Percentile Levels (default off): plots the thresholds as dotted levels.
  • Show Dashboard (default on): reading, benchmark, rank and half-life; Gradient Fill (default on) covers the fill.

Frequently Asked Questions

How does EWMA differ from historical volatility?

Historical Volatility weights every return in its window equally, so the line jolts when a big return enters and again when it drops out. The EWMA recursion has no window: a shock enters at full weight and fades continuously.

What does the half-life on the dashboard mean?

How many bars it takes for the weight on old data to halve, implied by lambda. At the 0.94 default that is roughly 11 bars, so the estimate is dominated by the most recent month of daily data.

Why rank the reading in percentiles?

Raw volatility levels do not compare across instruments. Ranking against the instrument's own lookback turns “high” and “low” into testable statements, and the alerts run on that rank.

Original indicatorBuilt in-house by LuxAlgo

The Library is free. Quant makes it yours.

Pull any concept or indicator into Quant: rebuild it, retune it, or turn it into a backtested strategy of your own.